Human Determinants of Motor Vehicle Driver Injury in a New Zealand Cohort Study Involving 108 741 Person-Years of Follow-Up

Abstract

Background Motor vehicle driver injury is a leading cause of injury, disability and premature death in New Zealand, but only limited data are available about the risks associated with several potentially important human determinants of driver injury. Objective The objective of this research was to investigate associations of driver injury with socioeconomic status (as indicated by educational level, occupational status and neighbourhood income), marital status, handedness, body weight and snoring status. Design A cohort study with prospective and retrospective outcomes Participants 10 525 participants were recruited from two sources: a nationwide multiindustry workforce (n = 8008) and a random sample of the Auckland metropolitan electoral rolls (n = 2517). Twenty eight percent of the study participants were females, 10% were Maori, 5% were Pacific Islands people, and 85% were European or other ethnicity. The ages at baseline ranged from 16 to 88 years (mean 44 years). Data Collection and Analysis Driver injury was defined as an injury received while driving a motor vehicle (e.g. a car or motorcycle) and severe enough to cause death or hospitalisation. Outcome data for the period from January 1988 to June 1998 were obtained from the New Zealand Health Information Service. Baseline data on exposures and confounders were collected in 1992-93. Data on educational level, occupation (coded using the International Socioeconomic Index), domicile address (used to identify median neighbourhood income), social marital status, handedness and snoring status were obtained at baseline by questionnaire. Body weight was measured at baseline by research assistants. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated by Cox proportion hazards regression. The HRs reported below were all adjusted for age and gender, except those for body weight, which were adjusted for gender and alcohol intake (there was little evidence of confounding by age). Results 139 driver injury cases were observed during 108 741 person-years of follow-up. Increased relative risks of driver injury were observed among participants in the lowest approximate quartile of occupational status (HR: 4.17; 95% CI: 2.31-7.55), participants who had attended secondary school for two to three years (HR: 2.15; 95% CI: 1.37-3.37) or for less than two years (HR: 2.26; 95% CI: 1.34-3.81), never-married participants (HR: 1.97; 95% CI: 1.29-3.01), participants who weighed less than 70 kilograms (HR: 2.24; 95% CI: 1.34-3.77), and participants who weighed more than 110 kilograms (HR: 2.18; 95% CI: 0.99-4.83). There was little evidence that driver mJury risk was associated with neighbourhood mcome, handedness or snoring status. Conclusions Occupational status, educational level and marital status appear to be determinants of driver injury. Body weight may also be a determinant of driver injury, but further research is required to corroborate or refute this hypothesis. This study did not have sufficient statistical power to exclude the possibility of weak associations with neighbourhood income, handedness or snoring status. The absence of an association with snoring status may have been partly or totally accounted for by misclassification of that variable. These findings identify high-risk groups to which driver injury countermeasures could potentially be targeted.